We show that wages in administrative data and in online markets exhibit considerable bunching at round numbers that cannot all be explained by rounding of responses in survey data. We consider two hypotheses—worker left-digit bias and employer optimization frictions—and derive tests to distinguish between the two.

Symmetry of the missing mass distribution around the round number suggests that optimization frictions are more important. We show that a more monopsonistic market requires less optimization frictions to rationalize the bunching in the data, and use this to derive bounds on employer market power. We provide experimental validation of these results from an online labor market, where rewards are also highly bunched at round numbers. By randomizing wages for an identical task, our online experiment provides an independent estimate of the extent of employer market power, and fails to find evidence of any discontinuity in the labor supply function as predicted by workers’ left-digit bias. Overall, the extent and form of round-number bunching suggests both employer mis-optimization and wage setting power are important features of the labour market.

About the Presenter

 Suresh Naidu is Professor of Economics and Public Affairs at Columbia University. He works in political economy, labor economics, and economic history. He has a Ph.D. in Economics from UC Berkeley, and an MA from UMass Amherst.